Some animals live together in stable groups while others prefer a solitary life style. But even the most asocial creatures may show a propensity to come together and aggregate under some environmental conditions. They may do so in different circumstances and for various reasons, yet all of them can be considered from a unifying evolutionary perspective. For instance, the ability to assess the environment and predict how it is going to change in the future is critical for animal survival and maximizing fitness. One way to obtain a more complete representation of the environment is to integrate information communicated by others. The genetically tractable nematode Caenorhabditis elegans, whose entire nervous consists of 302 neurons, is an attractive system for exploring neurogenetic mechanisms of animal decision-making in social contexts. The subject of this proposal is a recently discovered collective behavior of C. elegans L1 larvae-starvation- induced aggregation, which leads to improved survival. Preliminary results show that the aggregation requires at least two functional chemosensory neurons, called ASE and ASH, and is likely mediated by worm-derived chemical signals. We identified an enzyme essential for producing the aggregation signal-the alcohol dehy- drogenase SODH-1. Improved starvation survival of the larvae in clumps points to the functional importance of the aggregation for C. elegans fitness. Broadly, the goal of this project is to understand how neuronal, genetic, and molecular mechanisms in an individual contribute to an emergent population behavior, aggregation. At the individual level we want to know what signals the worm's exchange, what neurons in the worm respond to these signals, and what signaling pathways are activated in these neurons. These questions will be addressed experimentally with the tools of LC-MS and NMR chemical analysis and traditional molecular genetics. At the population level we want to know how animals come together and stay in clumps, what determines size and shape of these aggregates, and more importantly, what fitness benefits this behavior has for the worms. To this end we will computationally explore the aggregation using the active walker model, augmented to evaluate the contribution of aggregation to information accuracy and fitness. Mathematical modeling will be closely paralleled with experimental quantitative tracking of worm behavior. Specific aims of the proposal are - Aim 1: Determine neuronal and genetic mechanisms of starvation-induced L1 aggregation. Aim 2: Characterize the signals that mediate the L1 inter- actions. Aim 3: Develop a mathematical model for L1 aggregation and fitness. While the exact molecular mechanisms that operate in C. elegans may be species-specific, the role of social interactions in making critical decisions is a far broader question. Understanding the deep evolutionary roots of decision-making processes in social contexts, which in human society eventually lead to harmful or beneficial social behaviors, may help to improve our life style and mental health.